Many medications are small organic molecules that bind an enzyme or receptor that is involved in a disease process. Chemical hosts represent another class of compounds of biomedical interest; these are molecules which are much smaller than proteins but which still possess a cleft enabling them to bine1 a targeted ligand. Chemical hosts are used to improve the bioavailability and stability of medications, and have potential new uses as Pharmaceuticals in their own right; for examples as scavengers of toxins or even as artificial enzymes. Currently, there is no reliable computational method for designing targeted ligands or chemical hosts, so the discovery of such compounds relies heavily upon costly and time- consuming experimental trial and error. This proposal aims to speed the discovery of targeted compounds for medicalapplications by developing improved theoretical and computational methods, including ah automated approach to the design of chemical hosts, and an accurate new method of computing ligand-protein binding affinities. The approaches taken here focus on predominant states models, which have provided promising results in recent applications to host-guest systems in the Pi's laboratory. These methods are highly parallelizable, and this project will benefit from supercomputing expertise and resources available in the collaborating laboratory at the Mayo Clinic School of Medicine.